An Improved Music Composing Technique Based on Neural Network Model

Table 1

Evaluation results.

The name of the function	ur	Ustd	Ht	Ir	Gn	Jc	Dc	dxy	p
The original music	6	28.28	0.082	0	0	0.298	0.884	0	0.889
Exponential fitting	3	9.86	0.026	0	0.064	0.023	2	6.998	0.393
Exponential fitting	0	0	0.023	0	0	0	2	26.823	0.288
Gaussian fitting	33	69.368	0.228	0.298	0.298	0.068	0.828	60.02	-0.022
Polynomial fitting (N = 15)	8	23.393	0.082	0.039	0.42	0.039	0.896	2.802	0.426
Polynomial fitting (N = 20)	8	23.8	0.082	0.024	0.289	0.098	0.86	2.606	0.483
Polynomial fitting (N = 25)	8	24.008	0.082	0.282	0.269	0.23	0.883	2.408	0.824